National Institute of Technology, Silchar

About: National Institute of Technology, Silchar is a education organization based out in Silchar, Assam, India. It is known for research contribution in the topics: Control theory & Electric power system. The organization has 1934 authors who have published 4219 publications receiving 41149 citations. The organization is also known as: NIT Silchar.

A Cost-sensitive weighted Random Forest Technique for Credit Card Fraud Detection

Abstract: The act of fraudulent credit card transaction has been increased over the past recent years, as the era of digitization hits our day-to-day life, with people are getting more involved in online banking and online transaction system. Machine learning algorithms have played a significant role in detection of credit card frauds. However, the unbalanced nature of the real-life datasets causes the traditional classification algorithms to perform low in detection of credit card fraud. In this work, a cost-sensitive weighted random forest algorithm has been proposed for effective credit card fraud detection. A cost-function has been defined in the training phase of each tree, in bagging which emphasizes to assign more weight to the minority instances during training. The trees are ranked according to their predictive ability of the minority class instances. The proposed work has been compared with two existing random-forest based techniques for two binary credit card datasets. The efficiency of the model has been evaluated in terms G-mean, F-measure and AUC values. The experimental results have established the proficiency of the proposed model, than the existing ones.

Modeling, Simulation and Analysis of Surface Potential and Threshold Voltage: Application to High-K Material HfO2 Based FinFET

Abstract: In this study, an analytical model for surface potential and threshold voltage for undoped (or lightly) doped tri-gate Fin Field Effect Transistor (TG-FinFET) is proposed and validated using transistor computer aided design (TCAD) simulation. The threshold voltage with channel length 50 nm was compared with the published experimental results achieved from tri-gate FinFET. Separate solutions of 2D Poisson’s equation were obtained for both symmetric and asymmetric double gate FinFET and combined using the perimeter-weighted sum approach to achieve the surface potential for TG-FinFET. The inversion charge model was used to find the threshold voltage of the above mentioned device. The results of the model were obtained for different channel lengths, fin widths and fin heights on the silicon substrate. A comparative study of hafnium dioxide (HfO2) and silicon dioxide (SiO2) for the same oxide thickness was delineated to depict the influence of the high dielectric material on the model of FinFET. As anticipated, the oxide thickness of HfO2 is greater than SiO2 to maintain the same surface potential. The result of the analytical model was well agreed with the TCAD simulation outcome. Hence, it can be considered as a promising model in device technology.

Variability characterisation of input parameters in pavement performance evaluation

Abstract: Due to significant variability associated with the input parameters of pavement life and traffic repetitions, the performance evaluation in pavement design process becomes very uncertain. Uncertainty in performance evaluation can be handled using probabilistic design method. To calculate probability or reliability, it needs distributions and the parameters of distribution of the performance parameters, that is, pavement life and traffic repetitions. This paper attempts to establish these distributions and their goodness of fit based on statistical evidence. The uncertainty levels of traffic, pavement life and the pavement damage parameters are studied for the possible ranges of variations in different input parameters. Different possible combinations of distribution are considered, and it is observed that irrespective of the type of distribution of input parameters, the distributions of traffic, pavement life and damage parameters follow a lognormal distribution, with a wide range of coefficient of variat...